An image segmentation method based on Simple Linear Iterative Clustering and graph-based semi-supervised learning

摘要

Image segmentation is one of the most popular applications in contemporary computer vision and image processing field. In this paper, a novel segmentation framework is proposed based on Simple Linear Iterative Clustering (SLIC) and graph-based semi-supervised learning (graph-based SSL). With the SLIC approach utilized in the first stage, we get the anchor points which are SLIC clustering centers. Graph-based SSL method using anchor points and labeled samples are used to calculate category of anchor points, through which, the category of sample points is further determined. Experimental results show the effectiveness and robustness of the proposed method.